Sinica Leidensia Literary Forms of Argument

in Early China
Edited by

BarendJ. ter Haar Edited by

Maghiel van Crevel
Joachim Gentz and Dirk Meyer
In co-operation with

P.K. Bol, D.R. Knechtges, E.S. Rawsld,

W.L. Idema, H.T. Zurndorfer

VOLUME 123

BRILL
The titles published in this series are listed at bri!Lcom/sinl LEIDEN I BOSTON
 This volume received support from the "New Perspectives in Chinese Culture and Society" program,
which is made possible by a grant from the Chiang Ching-kuo Foundation for International Scholarly
Contents
Exchange to the American Council of Learned Societies.

Cover illustration: Wang]ie .=Effr (fl. 1331-1380), "Explanations and Commentary with Diagrams to the List of Contributors VII
Wondrous Canon of the Eternal Purity and Tranquility as taught by the Supreme Venerable Sovereign"
(Taishang Laojun shuo chang qingjing miaojing zuantujiezhu :;t.::J:~:!Et§~',ljt)j!f)Wfr'j>~~~~~fWtt), in
Introduction: Literary Forms of Argument in Early China 1
Zhengtong Daozang lE~ftm~ (Taipei: Xin wenfeng chuban gongsi, 1985-1988, vol. 28., p. 839), dongshen
bu, yujue lei, shizi hao )/lilt$ if~ · 35:§;1~~~ · ~'¥5ffE, (Dz 76o, fasc. 533, following the numbering in ]oachim Gentz and Dirk Meyer
Kristopher Schipper and Franciscus Verellen eds., The Taoist Canon: A Historical Companion to the
Daozang [Chicago: University of Chicago Press, 2004], vol. 2, 730 ). 1 A Building Block of Chinese Argumentation: Initial Fu :;k as a Phrase
Status Marker 37
Rudoif G. Wagner

2 Beyond Parallelism: A Rethinking of Patterns of Coordination and

Subordination in Chinese Expository Prose 67
Andrew H Plaks

3 On the Range and Performance of Laozi-Style Tetrasyllables 87

David Schaberg

This publication has been typeset in the multilingual "Brill" typeface. With over 5,100 characters covering 4 Defining Boundaries and Relations of Textual Units: Examples from the
Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. Literary Tool-Kit of Early Chinese Argumentation 112
For more information, please see www.brill.com/brill-typeface.
joachim Gentz ·
ISSN 0169-9563
ISBN 978-90-04-29160-7 (hardback) 5 The Philosophy of the Analytic Apergu 158
ISBN 978-90-04-29970-2 ( e-book) Christoph Harbsmeier
Copyright 2015 by Koninklljke Brill NY, Leiden, The Netherlands.
6 Speaking of Poetry: Pattern and Argument in the "Kongzi Shilun" 175
Koninklljke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi and
Hotei Publishing. Martin Kern
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, 7 Structure and Anti-Structure, Convention and Counter-Convention:
without prior written permission from the publisher.
Clues to the Exemplary Figure's (Fayan) Construction ofYang Xiong as
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive,
Classical Master 201
Suite 910, Danvers, MA 01923, USA. Michael Nylan
Fees are subject to change.

This book is printed on acid-free paper.

8 A Ragbag of Odds and Ends? Argument Structure and Philosophical
Coherence in Zhuangzi 26 243
WimDeReu
 CONTENTS

9 Truth Claim with no Claim to Truth: Text and Performance of the List of Contributors
"Qiushui" Chapter of the Zhuangzi 297
DirkMeyer
]oachim Gentz
Index 341
Chair of Chinese Philosophy and Religion, University of Edinburgh. Joachim
Gentz studied Sinology, Religious Studies and Philosophy in Berlin, Nanjing
and Heidelberg. He was Assistant Professor at Heidelberg (1999-2002),junior-
professor in Religious Studies at Gottingen (2002-2oo6), Visiting Professor in
Tokyo (2ooo) and Bayreuth (2oo8), and has worked at the Asian Studies De-
partment in Edinburgh since 2006. His main research focus is on Chinese his-
tory of thought. His work crosses the disciplinary boundaries of Sinology,
Religious Studies, Philosophy, and Cultural Studies. His recent publications in-
clude Keywords Re-Oriented (2009), Understanding Chinese Religions (2012),
and Religious Diversity in Chinese Thought (2013), ed. with P. Schmidt-Leukel.

Christoph Harbsmeier
Professor of Chinese, University of Oslo. Born in 1946, Christoph Harbsmeier
read Chinese at Merton College, Oxford, and was awarded a PhD from Copen-
hagen in 1981. He is a student of classical Chinese philosophy, grammar and
rhetoric with a special interest in modern Chinese illustrated literature. His
books include Wilhelm von Humboldt und die philosophische Grammatik des
Altchinesischen (1978), Aspects of Classical Chinese Syntax (1981), Socialist Real-
ism with a Buddhist Face: The Cartoonist Feng Zikai (1984), Language and Logic
(Science and Civilisation in China, vol. 7.1) 1998. Long-term engagements in-
clude: lecturer in Humanities, Princeton, visiting professor at University of
Chicago, University of California at Berkeley, University of Michigan, Chinese
University Hong Kong, Peking University, EHESS Paris, National University of
Singapore, Heidelberg, Wissenschaftskolleg Berlin, Collegium Budapest, Swed-
ish Collegium for Advanced Study, and the International Collegium Erlangen.

Martin Kern
Greg ('84) and]oanna (Pr3) Zeluck Professor in Asian Studies; Chair, Department
of East Asian Studies, Princeton University. Martin Kern is the author of The
Stele Inscriptions of Ch'in Shih-huang, the editor of Text and Ritual in Early Chi-
na, and, with Benjamin A. Elman, the co-editor of Statecraft and Classical
Learning: The Rituals of Zhou in East Asian History. He also is the co-editor of
T'oung Pao. His work cuts across the fields ofliterature, philology, history, reli-
gion, and art in ancient and early medieval China, with a primary focus on
poetry. He currently studies the formation of the early (Zhou through Han)
 GENTZ AND MEYER
YuanJinmei, "Analogical Propositions in Moist Texts."journal ofChinese Philosophy 39·3
(2012): 404-423.
Zach, Erwin Ritter von. "Einige Worte zu Prof. Gustav Schlegel's 'La loi du parallelisme
en style chinois'." Peking, 1902.
Zhang, Longxi. The Tao and the Logos: Literary Hermeneutics, East and l11est. Durham:
Duke University Press, 1992.
Zhou Yuanlong }Ell)J#!ij~, and Zhou Wei }Ell~, eds. Wenzi youxi X:'rw~. 2 vols. Beijing:
Tuanjie, 1999.
Zuckert, Catherine. "Commentary" to the contributions of Howland, Farness, and
Ausland to the volume Plato As Author: The Rhetoric ofPhilosophy, edited by Ann N.
Michelini. Leiden: Brill, 2003: 145-151.
CHAPTER 1
*
Building Block of Chinese Argumentation: Initial
as a Phrase Status Marker
Rudoif G. Wagner
In analytical non-narrative discourse, various rhetorical figures have been used
to mark the status of statements as either expressing general rules, exceptions,
or side comments. Such status markers might be silent and structural or ex-
plicit. Markers of both types exist in classical Chinese, but they are little stud-
ied. The consequence is a loss in the precision of the understanding of the
arguments proffered.
x
This is a study of the development of initialfo into a phrase status mark-
er for statements for which a general rather than particular validity is claimed.
The methodology used aims at falsifiability. After an outline of the state of the
art, the paper offers a qualitative analysis of a randomized sample of phrases
in Wang Bi's (AD 226-249) commentaries on the Zhouyi and the Laozi to de-
velop a plausibilized series of hypotheses. In a second quantitative step, these
hypotheses are tested against the entire body of these two texts. The last sec-
tion sketches the process by which initialfo shed its use as a demonstrative
and developed into the. structuring device for the Chinese rhetoric of argu-
ment outlined above.
Introduction
Language uses different markers to signal the status of a statement.1 A phrase
in English that starts with "basically" or "in principle" often indicates that a
statement is being made for which general validity is recognized from which,
however, in the actual case an exception might have to be made. The phrase
"thus I have heard" in the beginning of a Buddhist sutra marks the words that
follow as the words of the Buddha with all the accompanying claims to author-
ity this connotes. By contrast, a phrase starting with "let me add" or "by the
way" signals a statement that is supplementary to the core argument and will
1 Christoph Harbsmeier has been a helpful critic of an earlier version. Wolfgang Behr has kindly
helped with bibliographical references. The two editors as well as several readers have shared
their critical notes with me. To all of them I am greatly indebted. Needless to say, all mistakes
are my own.
 g8
WAGNER
not be further elaborated. These are explicit markers, supplemented in the
spoken language with sentence intonation, emphasis, and gesture. There are
also implicit markers. Examples are the closed grammar often observed in po-
litical or propaganda rhetoric2 or the position of a phrase within an argument,
such as a concluding phrase in a prose argument or a non-parallel concluding
phrase in Chinese interlocking parallel style constructions. 3 In many cases
these markers have been used as the basis for a modern graphic structuring of
texts originally written without breaks into phrases, paragraphs, and chapters.4
Using such structural as well as the explicit markers are economical and effec-
tive ways to structure an argument, and this even more so in texts written with-
out visible word, phrase and paragraph separators. A positive knowledge of
these markers is crucial for a precise understanding of argumentative proce-
dures in demanding texts from the 'foreign country' of the past, where so much
of the context originally shared by authors and readers is lost or as yet unex-
plored.
Chinese written status markers are treated by specialists under the negative
umbrella heading of xuzi @["'f=, meaning 'empty' characters. This only denotes
their not referring to a specific subject, action, or quality, but does not give a
positive definition. Scholarship has read this as an invitation to disregard
them.5
In an effort to enhance our understanding of the early Chinese rhetoric of
argumentation6 and, above all, the precision of our grasping the communica-
tions intended by the authors, I propose to study one word for which I claim
that it became a consistently used phrase status marker in argumentative texts
in the classical written language between the Warring States period (475-221
BC) and the third century AD. The word is.fo in the modern pronunciation (Me
bju < oc *[b ]a), and it is written with the character :X_. It very often occurs as
the first word-very rarely the second word-in a sentence, but also occurs
2 See Michael Schoenhals, Doing Things with Words in Chinese Politics: Five Studies (China
Research Monograph 41, Institute of Chinese Studies, University of California, Berkeley,
1976).
3 See RudolfWagner, The Craft of a Chinese Commentator: Wang Bi on the Laozi (Albany: suNY
Press, 2000 ), 53-114, for a detailed study of this type of style.
4 An example: Five of the eight paragraph breaks in Lou Yulie's edition ofWang Bi's :£385 Laozi
zhilile ~rf§IIJ& are marked by initialfo. Lou Yulie tf!''=pf,!{ ed., Wang Bijijiaoshi :-f385~
;f)(~l (Beijing: Zhonghua shuju, 1999), 1.195-99.
5 Examples would be the Thesaurus Linguae Sericae (TLs), accessible online under, http://tls.
uni-hd.de/ (accessed November n, 2014), or dictionaries such as Wang Haifen :£~*' Gu
Hanyu xuci cidian tl>Jifm:J.illl~BJ~iij~ (Beijing: Beijing Daxue chubanshe, 1996).
6 For a general and brilliantly comparative study of early Chinese rhetorical devices see
Christoph Harbsmeier, Chinese Rhetoric in Comparative Perspective, unpublished manuscript,
2002.
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU '*: 39
within sentences.7 A homoiophonous but probably different word.fo written
with the same character :X_ but a slightly different pronunciation (Me bju < oc
*ba) occurs only as the last character in a sentence. It indicates a rhetorical
question or a doubt, inviting assent ("is it not so?"). 8 The character :X_ is finally
used to represent the wordfii. (Me pju < oc *p(r)a), 'adult male.' 9 As particles
cann~t well be graphically represented, the use of :X_ to represent both initial
and terminal.fo must be assumed to be a phonetic loan based on the represen-
tation of the similar sounding (homoiophonous) fii. with the character :X_.10
The differentiation between the two homoiophonous fo ;X_, the one initial or
within the phrase, the other final, follows unequivocally from their position in
the phrase. The differentiation between these two andfii. ;X_, 'adult male,'-a
word that in addition hardly ever occurs either at the beginning or the end of a
phrase-is context-driven. It seems that these position and context differen-
tiations were unambiguous enough to obviate the need for further graphic dif-
ferentiation.
The State of the Art
Chinese commentators as well as, in their wake, modern scholars have noted
that initialfo occurs frequently, and that it often has something to do with
structuring text. Their common focus is on Warring States and Former Han
(202 BC-g AD) texts, their assumption being that the bandwidth of functions
and meaning to be found then is valid for the entirety of Chinese history. As a
consequence none of them traced the history of the word. The tone was set by
Liu Xie IUi~ (~465~522 ), who defined initialfo as "an opening salvo for a start"
7 In some rare cases, another word such as qie B or ruo ;fi may precede it, often without
causing a change of meaning.
8 I follow A.C. Graham in assuming that these two.fu are two different words and that.fu
'adult male' is a third one. See his "The Classical Chinese Topic-Marker fu :ft:," Bulletin of
the School of Oriental andA.frican Studies 35.1 (1972): Ss-no, 86. On the final particle.fu, see
A. C. Graham, "The Final Particle 'Fwu' :ft:," Bulletin of the School of Oriental and African
Studies 17.1 (1955): 120-132. B. Schindler's assumption that the fact that all these words
were written with the character :ft: was an indicator that they originally were one single
word that then split up into three seems not well supported. See his "Grammatical Notes
n: :;t(;• Asia Major, N.S. m.2 (1953): 162-168.
I follow the classification in the on-line supplementary list to William H. Baxter, Laurent
9
Sagart, Old Chinese: a new reconstruction (New York: Oxford University Press, 2014), The
Baxter-Sagart reconstruction of Old Chinese (Version 1.1, 20 September 2014), http://ocb
axtersagart.lsait.lsa.umich.edu/ accessed March 6, 2015.
10 Particles such as pronouns are as a rule written with loans. I am grateful to Christoph
Harbsmeier for pointing this out to me.
 40
WAGNER
(faduan zhi shouchang fNftfML §Of§) and grouped it with other words that start
phrases, such as wei H£, gai 'Ni., and gu i!&;ll by Lu Deming ~~{~Ej)j (556-627 ),
who marked initialfo as a "word used to start [a statement]" (faduan zhi zi ~if
ftfML ~)i and by Xing Bing lfG~ (932-1010 ), who observed the particular and
12
consistent function of initialfo in the Xiaojing ¥~.llf (Book offilial piety) and
defined it as "a marker for the beginning of a statement" (fayan zhi duan fjj:§'
3
Lftffii)l and in another passage he says it is "marking the start of a statement,
but it is also a demonstrative."14 In his Ma shi wentong ,liB!X,X::im (1898), Ma
Jianzhong .liB~~~' (1845-1900) quotes Xing Bing's statement as a source for ear-
lier definitions, but then developed his own. After grouping initialfo with three
other words of similar function, he writes: "generally speaking, iffo is at the
head of a sentence, it always serves to affirmatively link up with the text com-
ing before it (ding cheng shangwen J~71(J:X:) while in addition establishing a
new idea (chongli xinyi .m.TI:ifJr~) [ ... ]. Accordingly,fo is still a demonstrative
and not just an 'empty word' marking the beginning of a statement."I5 As many
chapters in philosophical works and even books of this type actually start with
initialfo, Ma's claim about the affirmative linkage to earlier text is not well sup-
ported. His effort to combine affirmative linkage, 'demonstrative,' and emphat-
ic aspects into a single ahistorical definition has set the framework for most
modern Chinese scholars. There are indeed early uses offo both at the begin-
ning and within sentences as demonstratives. An example from the Zuozhuan
will be given below.
Because the same character was used, scholars felt prompted to offer an
explanation covering both functions, even if these efforts remain awkward. In
1959 Zhou Fagao fE'J>t;~ classified initialfo as either being a noun akin to bi 1£!t.,
an adjective akin to na t1G, or a "demonstrative weakening into an initial
u LiuXie W{UmtJ, Wenxin diaolong )(,[,,f,lj~~~ chapter zhangju'1/i'ri), quoted in WangWeifeng
.:Efti~, "'Fu' de tongshe lianjie gongneng lun," ~i'l"JMt1Ji~:j:~:r;l;flliiinli, Gu Hanyu yaryiu
"d)Ji§filiJf~ 83 (2009.2): 63-68,63.
12 Lu Deming ~.i{i~J,BJ3,jingdian shiwen ~!Jt!!.V~:X:, quoted in Wang Weifeng, "Fu,'' 63.
13 And a few lines further down he summarizes: "to make clear that the reasoning from
before had its progression further down, therefore it says fu to begin it.'' Ming qian li er xia
you qi qu, guyan JU'yi qi zhi 8)3JW:£j[ffiff1'f:1'ti!!f:lt!N§~tiJe.:L. Xing Bing lf~~. sub-
commentary in Xiaojing zhushu ~~~ii±i&1L chapter 7, Ruan Yuan ~.71:51: ed., Shisarying
+
zhushu =~>±i&'lt (Beijing: Zhonghua shuju, 1987), 2558.1 (end). Xing right afterwards
also quotes Liu Huan JUJJI:, a philologist active during the Southern Qi, 479-502 AD, with
the words: "Fu is like fan ~j~fL -!:E."
14 Quoted from Xing Bing's commentary on Guo Pu's !)I~~ (276-324) preface to the Erya ~
!ffl; by Wang Weifeng, "Fu,'' 63.
15 Ma Jianzhong }i!'frll,'.t',, Ma shi wentong jiaozhu }il'fr~Y:i!Hi'3<'5±, Zhang Xichen ~~~~
comm. (Beijing: Zhonghua shuju, 1988), 355·
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU '*' 41
particle."l6 Inserting a historical dimension, he said that initialfo only gradu-
ally "transformed into something with the meaning of starting a statement and
even an 'empty word' of a purely statement-starting character, although the
borders between the two [the demonstrative, and the statement-starting,
R.W.] are not clearly marked:'17 This again seems less than convincing. Using
mostly quotes from the Mengzi, A.C. Graham maintained that initialfo was
referring only to the noun or nominal phrase directly following it, not to the
phrase as a whole, not to mention the pair of phrases following. 18 As we shall
see, this effort to link the demonstrative and statement-starting functions fails
to account for its very widespread use as a single marker shared by two parallel
sentences.
The recent compilations of specialized dictionaries for each of the thirteen
classics have not gone beyond claiming that the word "expresses a mood that
some summary outline" will follow. 19 In his study of the Huainanzi, Hans van
16 Zhou Fagao )j!!j$;i\1lj, Zhongguo gudai yufa tp~"d{-1;;§g$;, Chengdai pian ffl}{-1;;ffii,
(1959), repr. (Beijing: Zhonghua shudian, 1990 ), 177.
17 Ibid., 180, note 1.
18 Graham, "The Classical Chinese Topic-Marker,'' 92, 93· Strangely, Graham himself notes:
''A remarkable feature ofju, which has much to do with the impression that it is an initial
particle, is that it is not repeated in parallel sentences." Ibid., 94·
19 Many modern Chinese authors have defined, without further linguistic analysis, the func-
tions of initialfu. Pei X~ehai followed Ma]ianzhong in claiming an "emphatic function"
for it, 1'ft:lf::JGf!"ffl. Pei Xuehai ~~)#j, Gushu xuzi jishi "d:J!l!l["f=~V~ (Beijing:
Zhonghua shuju, 1954), 881. Taking the cue from Xing Bing, the Guoyu da cidian ~§fi:*:
~Jt!!. considers it a "speech opener" (fayu ci ~§fi~). Dongfang shudian *:1:1i!l;:l5
comp., Guoyu da cidian ~§fl:*:~Jt!!. (Taipei: Dongfang shudian, 1972) subfu ~.See also
Song Lian *)M (1310-1381), Pianhai leibian ffii )#j~Ji~i (Shanghai: Shanghai guji chuban-
she, 1995-99) renshilei ;\.$~,fu bu 7\:$. Luo Zhufeng ltlli'i-'f).ijR ed., Hanyu da cidian >Ji
§fi:*:iiiiJJt!!. (Shanghai: Shanghai cishu chubanshe, 1986-1994) sub fu 7\:, similarly makes it
into afaduan zhi zi ~fliffl.:L'¥. Xu Weijian ii'ff¥91 ed., Shanggu Hanyu cidian L "d>Ji§g
§ii]Jt!!. (Changchun: Jilin wenshi chubanshe, 1998), 98, calls it a character marking the
beginning of a paragraph, or a word which conveys a sense of the beginning of a sentence,
jushouyuqici 'ri)]'f§g~§B]. In the volumes that I have consulted (Mengzi, Lunyu,Xiaojing,
and Chunqiu Guliang ), initialfu is uniformly defined-with a formula originally proposed
by Yang Bojun ~{Bif&i: (1902-1992)-as a term "expressing a mood that some summary
outline" will follow ~t:lf:~§g. The expression tiqie t:lf:~ is the short form of the chengyu:
tigang qieling tf~~#J'i, "give a summary outline.'' See Shisanjing cidian bianzuan wei-
yuanhui +=~~!li!-~~~~~Wf, Shisanjing cidian +=~!~Jt!!. (Xi'an: Shaanxi ren-
min chubanshe, 2002) Mengzijuan, jfu-Toffi 76, Lunyujuan, iinli§fi;ffi 45, Xiaojingjuan ~
~;fir 13, Chunqiu Guliangjuan ;tf~.X~W:;ffi, 67. Wang Weifeng, "Fu," 63-68 reviews and
adds to earlier commentator statements on initial and demonstrativefu. His main contri-
bution is on thefu within phrases and its function to link elements together. Xu Kuangyi
§'f§;~ ed., tr., Huainanzi quanyi 5lJ¥ir~~~ (Guiyang: Guizhou renmin chubanshe,
42 WAGNER
Ess takes up the claim that the word indicates a phrase that takes up earlier
arguments and adds to them, often with the use of metaphor. However, none
of the examples given announces itself as a metaphor and there are several
cases where thefo starts a chapter.2 o
Commentators, as well as modern scholarship, have thus noted that initial
fo is a marker, but the efforts to define its function remain diffuse and have not
been tested with sufficient stringency. As a historian of Chinese philosophy
who believes that high accuracy in understanding the argument made in the
sources is crucial and who therefore has developed a philological bent, my
study is primarily concerned with the rhetorical function(s) which initialfo
assumes in structuring arguments in texts of interest to the student of Chinese
philosophy; the uses to which authors put initialfo; and the very practical im-
portance which a more accurate understanding of this word might have for the
study of Chinese philosophy. In other words, I am not primarily concerned
with mapping the bandwidth of the uses of the character 7( during the War-
ring States period, but with its role in a consolidated practice in argumentative
texts, if indeed it did develop such a role.
Consolidated Initialfo
During my studies of third century AD philosophical texts referred to as be-
longing to the Xuanxue 1':~ ('Scholarly exploration of the Dark') 21 current, I
have noticed a seemingly very special function which they give to initialfo,
namely to introduce general statements of principle. Two simple examples
from Wang Bi's .:£~8~ (226-249) "Zhouyi liieli" Jjlij~~{JU ("The Structure of the
Zhouyi"):
1995) vol. 1, p.2, note 1, offers a slightly improved variant in his note on the first initialfo
occurring in the Huainanzi, where he writes " :X_fo: "Mood particle for the beginning of a
phrase indicating that an argument will be made in the subsequent text." fi]l)'fj§fi*"§BJ •
oR7T;l'":X:~~~§All.
2o Hans van Ess, ''Argument and Persuasion in the First Chapter of Huainanzi and its Use of
Particles," Oriens Extremus 45 (2005/6): 255-270, 258. AB to initialfo being used for para-
graph segmentation, van Ess mostly refers to Western translations. The fine Huainanzi
edition by Xu Kuangyi quoted in the previous note, while not a critical edition, segments
the text. Of the nineteen segments into which Xu divides the first chapter, only eleven
start with initialfo, but the first chapter alone has 33 initialfo.
21 Rudolf Wagner, Language, Ontology and Political Philosophy in China. Wang Bi's Scholarly
Exploration ofthe Dark (Xuanxue) (Albany: SUNY Press, 2003).
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU '*' 43
Fu, the many cannot regulate the many, what regulates the many is that which is the
smallest in number.
The phrase explains the principle on which the relationship in many hexa-
grams between the single dominating line and the remaining five lines is built.
:J;(~Z?JT:>K=tt ~~m
~~Z?JT:>K=tf ~mzs
Fu, that what the yin is striving after is the yang and
that what the yang is striving after is the yin.
The phrase explains the basic dynamics dominating the relationship between
yin (broken) andyang (unbroken) lines in hexagrams.
In an attempt to solidify these hunches into a falsifiable argument with po-
tentially great practical applications, I will now set up and test the hypothesis
whether and under what conditions initialfo is used in 3rd century Xuanxue
texts as a status marker for argumentative statements claiming a general truth
and facticity; whether these statements were always of a specific formal type;
and whether the statements tended to be about a certain range of subject
matter.
Graham has nicely described the trap of selective examples. "Anybody who
insists on the liberty to select his own examples would be in a position to prove
any case whatever." 24 To avoid this trap I propose to follow a strict discipline of
randomization. This will come with a price because randomized examples
might need more complex elucidations of context. I will use the Wang Bi com-
mentaries and essays for a first test, because these are key texts in Xuanxue,
and I am most familiar with them.
I will proceed in two steps. First I will present the evidence at a stage where
the use of initialfo has consolidated. This will include the use of quantitative
methods and qualitative analysis to test the pervasiveness of a particular use of
22 Lou Yulie i'J'¥7.'!~ ed., Wang Bijijiaoshi :f.515~1'3(:f¥ (Beijing: Zhonghua shuju, 1999),
2.591. I will use this edition of Wang Bi's works on the Zhouyi. No critical edition of this
text has been made, and this edition is widely available. For Wang Bi's writings on the
Laozi I will use my own critical edition in Rudolf Wagner, A Chinese Reading of the
Daodfjing: Wang Bi's commentary on the Laozi with critical text and translation (Albany:
SUNY Press, 2003).
23 Lou Yulie, Wang Bijijiaoshi, 2.291.
24 Angus C. Graham, "The final particle fwu :J;(," 123.
 .44
the word in a given time horizon and the genre or type of statement where this
device is most commonly used. Second, I will shortly map the timeline along
which this use consolidated.
Two features set this approach apart from existing scholarship. First, it ac-
cepts the option of a historical trajectory in word or character use that may
include shrinking, diversification, and consolidation of meaning. This differs
from one element of the standard approach, namely that of trying to map the
totality of the different uses of a word without reference to historical change.
Second, it will try to establish a definition of the functions of this word that is
based on the analysis of the totality of uses in a given body of works, and a
random sample of other texts, and that thus comes with the claim to falsifi-
ability. This differs from another element of the standard approach, namely to
disregard quantitative patterns of use.
Fu appears altogether 49 times as the first word in sentences in Wang Bi's
Zhouyi commentary and 14 times in his systematic elucidation of the structure
of the Zhouyi, the "Zhouyi ltieli." In the latter text, all occurrences are in the first
25
part. In Wang Bi's Laozi Commentary, Laozi zhu :1;-T)£, initialfit appears 24
times, and in his "Laozi weizhi ltieli" :1;-T{J&:j:§~f§U 18 times with one occur-
rence only of fan }L. These are manageable quantities for a first test. As the
Zhouyi commentary does not set out to systematically introduce the uses of
initialfo, we may randomize our examples by simply taking the first few items
as they occur so as to test the general hypothesis and develop specifications. At
this stage the claim for their validity is tested only for the corpus outlined
above.
The first occurrence of initialfit is in Wang's commentary to the unbroken
line in the fifth position of hexagram 1, qian ~. All text in square brackets is
mine.
~! : 1L1iJ~:IJU±:7( 5fUJ%j(_A
)£ : ::ft=r::f~lffi:ft£-':¥:7( ~rmffiffPJ tJ:am:~J~ift :;;r~t~ 26 t£:7(~U*-AZ~~
=¥'1ft
:;K{_\1:~1,~-
i~~{f[~
.W~1~fflfl±\~{ft lttfo/J:L:ft ::f?J)\'§:f
25 In the second part, the "Liieli xia" ll!g{9fT, the termfit does not appear at all in this posi-
tion, but the term fan fL, which by this time had acquired a similar meaning, occurs five
times in the initial position otherwise occupied byfit. I have as yet been unable to account
for this shift in usage.
26 I read 1~ for 1-i in accordance with Guo Jing (d. 1127) ¥~*' Zhouyijuzheng }lli]J!)Hf!JE,
quoted in Lou Yulie, Wang Bijijiaoshi, 1.220, n. 18.
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU '!< 45
The text: Unbroken line in the fifth position: Aflight the dragon rests in the sky:
the [others] will have the benefit of seeing the Great Man.
Wang Bi Commentary: "As he [the dragon] does not act and does not jump [any-
more as he did on the two previous lines] but 'rests in the sky' how else would
this be done but by 'being aflight?' That is why [the text says]: 'Aflight the dragon.'
Once the dragon has achieved it to 'be aflight in the sky;' the path of the Great
Man will be successful.
Fu,
a position will be brought to full fruition through receipt/capacity; receipt/
capacity will provide order by means of a position. 27
As through supreme receipt/capacity he is residing in the ultimate position, is
not the 'observance' [towards him] by the 'ten thousand kinds of entities' [of
which the Master in the "Wenyan" commentary speaks] appropriate [=to be
expected] ?"28
I have arrived at my reading of the Zhouyi passage from Wang Bi's commentary
through extrapolative translation. Wang Bi's commentary starts out with a spe-
cific analysis of the Zhouyi phrase. He then inserts after fo a general comment:
"A position will be brought to full fruition through receipt/capacity; receipt/
capacity will provide order by means of a position." The two parts of this gen-
eral comment are parallel. They have the same number of characters, the same
grammatical structure, and analogous characters are in the same position.
They are dialectically linked and not just juxtaposed separate statements. This
leads to a first specification of the hypothesis:
27 The term 'position' here refers to the social rank as symbolized by lines two to five in a
hexagram. The fifth position, to which this commentary refers, is that of the ruler.
'Receipt/capacity' refers to the double meaning of the word de (which is often translated
as 'virtue') as being a 'capacity' f,W, 'received' 1~ from the dao. It is through this receipt/
capacity that the potential given with a royal position will be brought to fruition.
28 Lou Yulie, Wang Bijijiaoshi, 1.2IZ. The "Wenyan" commentary, to which Wang implicitly
refers, runs: "What does 'Aflight the dragon rests in the sky; the [others] will have the
benefit of seeing the Great Man' mean? The Master said: 'Corresponding sounds echo
each other, similar qi attract each other. Water flows towards the moist, fire goes after the
dry, clouds follow in the wake of the dragon, wind follows in the wake of the tiger. When
the Sage fulfils his function [as ruler] and the ten thousand kinds of entities are in obser-
vance, those rooted in heaven will be close to him above while those rooted in earth will
be close to those below so that each follows its own kind."''¥ El fbl~113JJ.i fbl ~i'§:>JZ 7]<.
)Jit)~ ;J(Jltt#~ ~1fl':~g )@,1fl':df.. ~A.1'Flffli!l?l0m 2Js:'fx:ff~JLI: 2Js:'f:t:ili:ff~Jrf ¥!U%
1fl':~!J!ffi!, Ibid., 215. I cannot follow Lou Yulie who suggests to read the zhi Z in ~!fo/JZ'lm
as wang 11 'to go to.' Ibid. 220 n. 20. There is no basis for this in either the Zhouyi text or
the "Wenyan." Richard Lynn, I Ching (New York: Columbia University Press, 1994), 137,
does not translate the.fit.

(1) Initial fu is not simply a speech, or paragraph, opener and does not contain a
demonstrative element. It has the JUnction of marking the status of the sentence
which it introduces as being ofa higher validity than sentences not marked in this
manner. Such a sentence might consist of two interlocked parallel segments. 29
The fo phrase spells out a basic organizing principle of the hexagrams in the
Zhouyi that has an evident bearing on political life. While lines one and six at
the extremes mark the beginning and end of a process, lines 2 to 5 each have a
'position' wei {iz: with structural links to other positions. Line five, for example,
is the position of the ruler. Depending on the hexagram, a particular type of
person with a given 'receipt/capacity' (de 1,~) is in such a position. This person
is symbolized by a line. The basic classification is into 'weak' and 'strong' re-
ceipt/ capacity as symbolized by a broken or an unbroken line on the given
position. The position of an emperor is elevated, and his actions are consid-
ered the root cause both of order and of disorder in society. Wang Bi and others
have written much about this. Only if this high position is coupled with de
f,~-and Wang Bi elsewhere defined this as de 1~, 'get, receive' in the meaning
of what is "received from the Dao"-will it unfold its positive effects. Someone
with the de f,~ of a Sage, but without high position, as was the case with Con-
fucius, will not be able to bring order to society. This will be possible only when
coupled with a high position as in the ideal of the sheng wang ~_:f., the sage
kings of old.
The fo phrase is thus linking the hexagram coding in a highly abstract, gen-
eral, and philosophical way to a proposition in political philosophy that is
shared by a wide array of relevant writings and examples. The claim of the fo
phrases to validity is further enhanced by the fact that they are often followed
by agu ti)t 'that is why. ..' as will be documented further down. From this follows
a second specification:
(2 ~ I~itial fu is a rhetorical marker signalling to the reader the statement of a
pnnctple or a rule underlying a body of statements and structures in canonical
texts that are supposed to be known to the reader.
The sentence followingfo does not tell a story but presents a non-narrative
statement of argument. It is articulated in a highly abstract and general way.
This gives us a third specification:
29 Graham wrongly claimed that initialfo only referred to the subsequent noun or nominal
phrase.
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU"' 47
(3) Initial fu is followed by a non-narrative statement of argument or principle
with abstract vocabulmy.
Thefo sentence does not quote a canonical text but occurs in a commentary to
a canonical text. The implicit claim is that without the content spelled out in
the fo sentence the canonical text remains incomprehensible and that there-
fore thefo phrase spells out as a new abstract philosophical proposition what
in an implicit way is contained in the canonical text. It does so, however, in an
inter-communicatively rational manner that does not hinge on authority.
The canonical texts derive their authority not from being revealed 'scrip-
ture' but from being the bequests of the sages of antiquity. These sages are de-
fined as a small number of human beings with an inborn grasp of the complex
workings of the universe, and its ficlde brother, human society. 30 Nothing re-
mains of the wealth of examples from which this rule has been distilled. The
historical concreteness, and anecdotal specificity, the unending flurry of par-
ticular historical circumstances and textual utterances have all been replaced
by a sober, rigidly rational rule about the dialectics prevailing between receipt/
capacity and position. To extract this rule from the concrete material at hand,
and insert it to make people understand the logic of a canonical text is Wang
Bi's own contribution, and this is in fact the translation of the Zhouyi text alto-
gether from a tool for divination into a philosophically grounded analysis of
the political process. Thefo thus refers to propositions spelling out the princi-
ples underlying communications of 'higher validity' such as those addressed in
the classics rather than just trivial regularities. Buddhist authors would use this
rhetorical device to refer to truths underlying Buddhist texts; the Huainanzi
will expand its canonical authorities to other persons and to processes in na-
ture.
Rhetorically, the fo phrase here states, in a most general way, a rule. Wang Bi
inserts his general key for understanding the Zhouyi coding for position (wei
{iz:) and de 1,~ right into the middle of the argument. The statement comes
without further proof, and it has no precedent in the language of the "Wenyan,"
which, being associated with Confucius, has the stature of a commentary from
the hand of a Sage. The fo phrase is not accidental to the argument. The logic
of the argument actually hinges on the acceptance of the fo sentence. At the
same time, the insertion of the fo sentence establishes the link between the
Zhouyi language and structure on the one hand and the "Wenyan" interpreta-
tion on the other, which is quoted in the succeeding sentence, and is thus prov-
en to be true.
30 As I have tried to show elsewhere, Wang Bi does neither accept a proof by authority, nor
does he set himself up as a teacher whose word has to be accepted even without proof.
Wagner, Language, 78.

(4) Initial fu alerts the reader to statements claiming to spell out a new insight in
an inter-communicatively rational and argumentative way about the general
rules or principles underlying texts or reports ofhigher truth, proving their philo-
sophical viability on the one hand, and affirming the appropriateness of their
high status on the other. It is not a markerfor a statement claiming the unques-
tionable authority of the classics as a proofof truth, but comes in a Larger context
that provides the evidencefor the access of the classics to truth.
The analysis of one example netted a fourfold specification of the hypothesis.
To verifY the plausibility of the hypotheses and the specifications and perhaps
add others, I will now go through the next examples as they occur in Wang Bi's
commentary.
Text: Unbroken line in the top position: An overpowering dragon will have cause
of regret. [But] when making [proper] use of the unbroken line, he will be show-
ing it to the many dragons. Not being their head [=overpowering leader] is where
fortune lies.
Wang's Commentary: "The unbroken (yang) line signals the receipt/capacity of
heaven. Being able to make [proper] use of this heavenly receipt/capacity has
the meaning of 'showing [the right way] to the many dragons.'
Fu,
if [a ruler] makes use ofrigour and strength [that is, qian lf!i;], but has his place
as the head [=as the overpowering leader] over others, he will be one with whom
the others will not associate;
if [a ruler] makes use of softness and obedience [that is, kun ;hfl], but acts as a
Lord who is not orthodox, this is the way of [bringing about] viciousness and
heterodoxy [among those below].
That is why the 'fortunateness' of qian is in 'not acting as the head,' and the
'benefit' of kun is in being 'perpetually orthodox."'32
31 The characters zhi zhu ~± are added on the basis of the Guben E2fs: and Zuliben
JE;fU2fs: manuscript quoted in Ruan Yuan ~}tj[;, Zhouyijiaokanji fi'D~iY:WJiic, in Ruan
Yuan ed., Shisan jing zhushu: fu jiaokan ji +.=:.~!).!ilfrt: IWiY:Wriic (Beijing: Zhonghua
shuju, 1987), 1.gb.
Lou Yulie, Wang Bijijiaoshi, 1.212-213. The statement about kun here quotes from a phrase
m
about the top line in the kun hexagram. It runs: 1\ ;fUli<.~. "making use of the broken
line will have benefits [for the ruler only] if he is perpetually orthodox." Hexagram kun is
entirely composed of broken lines. Lou Yulie, Wang Bijijiaoshi, 1.228.
.A BUILDING BLOCK OF CHINESE ARGUMENT.ATION: INITIAL FU "*': 49
Thefo sentence in this example precedes a statement similar to that in the first
example. It is followed by two parallel phrases, both of which end this time
with the copula ye fu, The initialfo again cannot be attributed only to the first
of the two phrases, but pertains to the statement as a whole. The fo phrase is
non-narrative and argues a general principle. It exemplifies it with two anto-
nymic cases, the ruler who uses the features of the qian hexagram, namely rig-
our and strength, but then exaggerates this feature by lording it over the others;
and the ruler who uses the soft features of the kun hexagram, but then exagger-
ates the 'softness' of these features by being cunning, sly, and 'not correct.' Both
deviations end up being counterproductive, the ruler following the first has no
officials who will support him to bring order, and the ruler following the sec-
ond deforms the bureaucracy and society through his model. That is the "cause
for regret" of the Zhouyi text. Thefo phrase here brings together parallel state-
ments in both the qian and the kun hexagrams and tersely shows in rigid paral-
lelism the philosophical principle underlying the canonical narrative, which in
itself seems vague.
The philosophical character of the fo phrase here is evident from it being
counterintuitive. Of course, the ruler who uses "rigour and strength" will be
expected to be a strong boss over the others, and the ruler who rules by being
"soft and accommodating" will be expected to have cunning ways of getting his
will. A highly rational and counterintuitive principle of political philosophy is
extracted from statements in two parallel hexagrams which in themselves are
rather cryptic. Thefo sentence shows its function as the philosophical explana-
tion providing the bridge between the first part of the text that is commented
upon (yongjiujian qunlong ffl1L J!f:Fij~) with the second (wushouji ~1i a)
by being continued with a 'that is why' gu r£;)1:. This gu confirms the notion that
what was previously stated was a principle or a general rule, as it is now ap-
plied to a specific textual passage. What comes after this gu is not the new in-
sight, but a recapitulation of the text itself, which is now to be re-read in the
new light provided by the precedingfo sentence. For this reason gu should not
be translated as 'therefore' but as 'that is why [the text says ].'33 This means that
only by way of thefo phrase will the reader be able to understand the underly-
ing reason for the two counterintuitive statements about the qian and kun
Gu ii)( is another literary device to structure an argument that has not been properly ana-
33
lyzed. In the textual environment of Xuanxue, it is followed by something familiar to the
reader-often, as in this case, a verbatim quotation from the text Wang Bi comments on.
The new argument or principle precedes the gu rather than follow it. The reader is
expected to revisit a familiar phrase or proposition in the light of the new interpretive
framework given before the gu. It therefore has an emphatic character: "This [and not
what other commentators might have told you] is the reason for ... [the text saying xyz] !"
See Wagner, Craft, 264-265.

'so
hexagrams. In terms of political philosophy, the argument ties in closely with
Wang Bi's arguments in the Laozi commentary.
The passage confirms hypothesis 1 by being a rhetorical marker; hypothesis
2 by spelling out the philosophical basis for a canonical text; hypothesis 3 by
being non-narrative; and hypothesis 4 by spelling out in a rational way the
principle underlying a canonical text. While the resulting political philosophy
might be one that is closely associated with Wang Bi, its insertion into the
Zhouyi as providing the key to the understanding of its cryptic utterances
makes at least a claim that this political philosophy is not the predilection of
just a school or even a single person, but the underlying thinking of the clas-
sics. This twofold status enhancement, of the classic as a philosophically
grounded text, and ofWang Bi's philosophy as a key to unlock the teaching of
the Sages, has the intended effect of not just appealing to a reader's common
sense or previously shared opinions, but of convincing him of the correctness
of this philosophy through the fit between the statements in thefo phrases and
the structural pattern dominating this canonical text. The success ofWang Bi's
reading in later Chinese history should be kept in mind in this context.
This might be the moment to suggest a way to make the content of initialfo
explicit. I suggest going from the non-translation of initial fo as an 'empty'
character all the way to the opposite end by making very explicit what is im-
plied in the use of this word. The reason for such a radicalism is that for the
reader of a translation-and every modern reader, including the scholar with
Chinese as his or her native language as well as the Japanese or Western sinolo-
gist, is understanding these early Chinese texts through an implicit or explicit
translation into a modern vernacular-the particular functions of initialfo
have to be boldly marked and highlighted to help the reader understand the
way the argument is made, to guide an informed reading strategy, and to
achieve optimal precision of understanding.
My cumbersome suggestion would be that initialfo is a status marker for
statements of principle that might appropriately be translated as 'it is a general
rule or principle that....' Needless to say, this verbalization of the functions of a
rhetorical marker is just an attempt to express its functions in detail. In any
given specific situation, other, and shorter, forms might be found to signal the
status of the ensuing phrase. For the two examples hitherto studied this would
give "[i]t is a general principle that a position will be brought to foil fruition
through receipt/capacity, and that receipt/capacity willprovide order by means
of a position," and "[i]t is a general rule that if[a ruler]makes use of rigour and
strength [that is, qian J¥Z;], but has his place as the head [=as the boss] over others,
he will be one with whom the others will not associate; if [a ruler] makes use of
softness and obedience [that is, kun :1:$], but plays the lord who is not orthodox,
this is the way of [bringing about] viciousness and heterodoxy [among those be-
low]."
p. BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU 'I< 51
The third example explains the first phrase in the tuan %c statement for the
ian ¥Z; hexagram, :k~i¥Zn ;§t~J;ff<E1 JJMf.':R., "Great indeed is qian in its be-
ing the origin! The ten thousand kinds of entities draw on it for their beginning
and thus it encompasses heaven." The commentary starts:
'7:. ttz,:=g ID:L ::5 ili
¥Z34 ttz,:=g miD::gttz,
~% -tl!.:!!f !ftz~-m
1§'7:.:LID fffl~11<15f!d*!l5
t.w~:L§ ~)t;L;g
_R~p~{?!~
Heaven [while being great] is [still] the denomination of a form [insofar as it is
restricted to covering all, but not able to support all, which is what earth does].
Qian is that which is making use of this form [of heaven]. It is a general princi-
ple that form is that which binds entities [to particularity]. To have the [lim-
ited] form of heaven, but still be able to be preserved eternally without decay,
and as the head over the entities be that which holds them together, how could
35
this be anything but [as the 'Xici' says] 'the highest strength'?!
Thisfo phrase again ends with aye tt?,. The insertion of this phrase is of crucial
importance for explaining the exclamation at the beginning. The first two def-
initions on heaven and strength also have the definitional ttz,:=g but they stay
within the language and· particulars of the Zhouyi. Thefo phrase alone articu-
lates a general principle. It defines entities as being bound and limited by form,
even entities of the 'great' kind such as heaven. Limited by form also means
being subject to the vagaries of time. The inserted phrase does not follow from
the immediate context, and is not an echo of a statement somewhere else in
the Zhouyi ~r its wings. At the same time, it is a definition that appeals to im-
mediate acceptance. Once the argument is set up that even heaven is the name
34
:rz: for ft! on the basis of the sub-commentary by Kong Yingda, as first suggested by Guo
Jing.
Lou Yulie, Wang Bijijiaoshi, 1.213. The last phrase does not seem well transmitted. ~to/JL.
35
tff ~!t:L. ~ does not hold well grammatically. As the text just quoted above made it clear
that wu shou 11\iitfr "not acting up as the head" brings good fortune,jis,it is not clear why
here 'being the head' should be extolled. I assume the text might have read in roughly
parallel fashion 1'f:7(L.JfJlffiflbJ<.fif;1!\iil§ ::f~to/JL.tff lffi~L.~ ~~~¥fl!u~. "To
have the [limited) form of heaven, but still be able to be preserved eternally without
decay, not to be acting as the head over the entities, but still be that which holds them
together, how could this be anything but the highest strength?!" My reading of thefo
phrase does not hinge on this conjecture. The ¥fl! takes up a statement in the "Xici;'
;:}i;: :$Z:7( r:L.¥fl!fu "Generally it is true that qian is the highest strength under the sky:'
Lou Yulie, Wang Bijijiaoshi, 2.537·
 52
WAGNER A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU 12:
53
of something that has a form, the general statement that forms are what limits receipt/ capacity; truthfulness that he settles [his] task.
and binds entities presents no challenge. However, for the subsequent rhetori- Insofar as he
cal question it is only through this insertion .that the description of the highest understands getting to [the beginning] understands bringing things to a
or ultimate strength that is beyond big and small makes sense. The da zai j(~ and brings about this getting to, it is conclusion, and brings them to this
is then linked to the zhijian ~{I in the "Xici," and both signal something of an possible for him to join in getting near conclusion, it is possible for him to
absolute order that makes use of forms, but is not bound by them. [to duties being achieved]. join in preserving righteousness.
Again it should be noted that the theme of name and form, and that of the
nameless and formless is a very specific Xuanxue topic that is much explored Wang Bi Commentary:
[This line],
by Wang Bi. Still, the fo phrase cannot be said to make an argument through a
is situated [below] the [lower] extreme is positioned at the end point of the
backdoor imposition. While there is no hard proof here that fo might not be a of the [upper] trigram, this is what [the [lower] trigram, this is [what the text
topical marker referring only to xing ~~ the main point of the phrase is in the text refers to with] "getting to [the refers to with] "conclusion."
second part, and our hypotheses are all supported by this example. beginning]."
The fourth example occurs in Wang Bi's comments on the "Wenyan" :X§ Being situated at the starting point of Being situated at the end [of a
commentary to the third unbroken line counting from the bottom. process],
a process and not committing something and being able to keep it fully intact
:X§ : JL-=: El :!Sr~~ B ¥l:¥Z: Y''f~1ilfj ~~ f6J"iJ,IlH±?, that will have to be regretted is [what to its end is [what the text means
rEJ the text means by] "[he] understands by] "[he] understands bringing things
;s-=r getting to [the beginning]." to a conclusion.''
~t\\!! That is why [the person occupying this Lacuna. 38
,~,{§ pff~~:t\\!!-tt line] is "able to join" in getting
9ill~~~ PJW~tt assignments done.

5± : As a general principle it is afact that

b1t-Jil~i'~ ~~tt @-3:'~ ~~ ~~~tt
as far as the speed of bringing things as far as keeping things together to
b1t$~~ffiJ:::fJ8~ 9Ql~:tftt forward is concerned, righteousness their end is concerned, profit does
b1t~~rmti5~ 37 ~~~ ~o~~:tftt cannot compare to profit. not even get close to righteousness.
tlti:PJW~r%% [lacuna]
;fr; That is why [as the "Daya" of the Book of Songs says]
"there is nothing that does not have a
;lf§!Jey 21!i'f *lfftf~tl
tli)(: beginning [but] few [things] manage to have a proper
B/F1HJJ conclusion.''39
;fr; PJ!li!#.i'f ~ Pflf~ ~i'f-'¥
As a matter ofgeneral principle it is afact that
only the one for whom it "is possible to join in
"Wenyan": "What does it mean that the [text accompanying the] unbroken line preserving righteousness" is the person who
in the third position says 'If the gentleman will be active all day and even in the "understands bringing things to their conclusion."40
evening he still takes care, he will suffer no blame'?" The master said:

"The gentleman
advances [his] receipt/ capacity
It is honesty and trustworthiness
through which he advances [his]
36 There is a probable transmission mistake here. Either it should be f~* in both cases or
We have twofo sentences here. Neither one ends with aye -!:t. The first is fol-
lowed by a pair of parallel segments, to both of which it pertains. The first
and cultivates [his] task.
and it is through carefully choosing
38 There is a lacuna in the text here that parallels ii)(11JWJ7JG:r%~ of the phrase about ~0~.
[his] words and establishing his
It should run about as follows: ti\(PJ Wf¥~~ or offer a "translation" of ff~. This is
confirmed by the parallelism in the "Wenyan" text. I also assume that )]!&~~ should be PI&
~z~~.

37 '**·
~ instead of~ based on one MS quoted in the Siku quanshu edition.
39
40
Maoshi =§8i'j, no. 245·
Lou Yulie, Wang Bijijiaoshi, 1.214.
 54 WAGNER
segment provides a principle underlying a factual statement in the "Wenyan."
~or this purpose it takes up one of the terms offered by this text, namely yi ~~
nghteousn;ss, ~nd contrasts it with another term not offered there but 'sup-
pl~mente~ as Its regular antonym, namely li fU, profit, to arrive at a general
philosophical statement concerning the fundaments of effective state man-
agement. The antonymic pairing of li andyi is familiar from many discussions
about state management. The sentence confirms all of our four hypotheses.
The concludingfo sentence drives home the lesson by abandoning the parallel
style symmetry, combining in this way two status markers, a structural one that
ma~ks the phrase as a general statement pertaining to both of the parallel
chams, and a second one through the initialfo. It picks up and explains the
imbalance between the parallel statements about zhi ~ and zhong ~t in the
"W~nyan," which is sig~alled by the fact that only zhong is adorned with a high
register term, namely nghteousness, and drives home the point in the form of
an abstract principle. Again our hypotheses apply.
We conclude this round with the last two examples from hexagram 1. In the
comment to the "Wenyan" about the unbroken line at the top of the first hexa-
gram, Wang Bi writes in a mock dialogue form that patterns itself on the "Wen-
yan":
l
The 'We~yan' on .[hexagram qian initially does not talk about qian, but only
after haVIng explamedyuan 7C [and the other three, heng Jjt, li fU and zhen J;l'[]
it speaks about qian. Why?
"As a ma~~r ofgeneral principle, qian is that which overarchingly acts out these
four quahtres [of yuan, heng, li, and zhen ]. It is because [as the xiang ~ com-
ment~ry to this ~exagr~m says] 'the gentleman will by way of strengthening him-
self Wit~out rest practice these four that initially [the "Wenyan"] does not talk
about qtan, but [only] further down says 'qian is [=encompasses] yuan, heng, li,
and zhen."'41 . section of his "Zhouyi liieli" (ZYLL ). 43
Thefo phrase here offers a definition of qian that is well documented and con-
firms the rational correctness of the "Wenyan" explanation. Again our hypoth-
eses are supported.
The last example again offers a mock question-and-answer sequence.
t%3<. ~li>tJJ~~~n-=:.1!lllPJttr~§ fPJili
41 Ibid., 1.215.
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU *: 55
;X. J'bi!f~.W. ~Z.?JT1: 1:n~~ili *FWT~f~1&E1JlZ.PJ~!Jo/J t)ZPJJ3~#0Z~z PJ
,i,BI3f3f$ ilJI~$~ffiJ!&~J.% ~i!i)Z:fJJJL)L= JJ~f!(~~~~ i!i)(BJ§ffi!Jj~tzE!Jl
:Lift ~~11= ·~Mz:>t·tJb' ~r~1!\ili 13J3PJttr~~~~ ~3trm~z. ~ill
~JJ~ J:JUffiJ~S/:Z. ~llJI~~ili
The remaining lines [in hexagram qian] are all explained [by the line commen-
taries] with reference to the dragon. When it gets to the unbroken line in the
third position [the line commentary] makes an exception and makes the gentle-
man the topic. Why?
·~s a matter of general principle it is a fact that the [Zhou ]yi is [=consists of]
images. What generates the images lives from the meaning [intended by the
Sages]. Once [they] had a given meaning, they would illustrate it through the
appropriate object. That is the reason why they use the dragon to sketch [the
meaning of] qian, and the horse to illustrate [the meaning of] kun, taking images
according to the meaning of the matter under consideration. In this sense, for
the first and second unbroken lines the receipt/ capacity of the dragon corre-
sponded to the meaning-and that is why it was possible to talk about the
dragon to explain their [meaning]. Getting to the third unbroken line, the 'active,
active, evenings [still] worried' is not [part of] the receipt/capacity of the dragon
and illustrating it by way of the gentleman fits this image. Altogether, the sub-
stance of qian is dragon throughout, but when deviations are detailed, they each
correspond to a particular meaning:•4z
This radical definition transforms the Zhouyi from an oracle handbook into a
philosophical text. As the meaning intended by the Sages defied definitional
language, the images serve as a pointer beyond themselves to this meaning. By
defining the Zhouyi as images, Wang Bi took up much of the structure of the
Zhouyi itself with its xiang commentaries as well as statements in the "Xici"
about the way in which the Zhouyi had been created by the Sages. At the same
time, this definition offered the foundation for Wang Bi's own interpretive
strategy, which has come to dominate the Chinese interpretive tradition. Wang
Bi further elaborated on this in the "Ming xiang" :21~ (Explaining the images)
In a second step, I will now move to a more quantitative approach to define
the types of sentences following after initialfo. The data will be the 49fo sen-
tences in Wang Bi's Zhouyi commentary44 and 14fo sentences in his ZYLL. 45 Of
the 6gfo sentences of the sample, 27 (Ig/8) are followed by parallel segments
42 Lou Yulie, Wang Bijijiaoshi, 1.215-216.
43 Lou Yulie, Wang Bijijiaoshi, 2.6og.
44 Inhexagrams1(7),2(2), 3(2), 4(1), 8(5), g(6), 12(1),14-16,26,27, 32, 34, 42,44-45,47,48 (one
each), 49(2), 51(1), 52(1), 54(1), 56(2), 57(1), 62(1), 63(3), 64(1).
45 "Ming tuan'' 1¥J~ (5), "Ming yao" i¥JX (z), "Ming gua'' 1¥Jtr (g), "Ming xiang" 1¥J~ (1),
"Bian wei" ¥iJ¥{:fL (z), "Gua hle" trffi& (1).

with the single initialfo pertaining to the entire pair.46 z8(z6/z) make an ana-
lytical statement of the ldnd "As a matter of general principle it is a fact that
only the one for whom it "is possible to join in preserving righteousness" is the
person who "understands bringing things to their conclusion;"47 and eight give
a definition of the kind "it is a matter of general principle that form is that
which binds entities [to particularity] ."48
Under these circumstances it seems that the traditional notion of initialfo
as a topical marker is not further helpful for our analysis. Initialfo is used here
throughout as a status marker for statements and definitions of a general phil-
osophical nature. Understanding its function is in fact crucial to understand
the argumentative strategy and result. The phrases throughout confirm the hy-
pothesis and its specifications as outlined above. In the ZYLL, thefo phrases are
not inserted into a commentarial environment, but offer the underpinnings of
Wang Bi's own argument. Within the Zhouyi Commentary, they are inserted as
key philosophical underpinnings of Zhouyi utterances that themselves have no
visible philosophical content or connection. Of the 49fo phrases in the Zhouyi
commentary, 15 are followed by a 'that is why.'49 In the ZYLL the 'that is why'
normally does not refer to a verbatim formula used in the main text, but rather
to a pervasive structure of the main text. Three of the 14 items have such agu.so
46 In Wang's comments to Hex. 1(g), 2(2), g(1), 4(1), 8(1), 9(1), 14(1), 27(1), g2(1), 42(1), 4g(1),
47(2), 49(1), 56(1), 6g(1). In his ZYLL "Ming tuan'' (g), "Ming gua" (g), "Mingxiang" (1), "Gua
Hie" (1). There is one case in the first section of the ZYLL with two subsequent parallel
phrases, each one starting with afo. :;K ~::ftl55€:l~ 5€:;~~¥-~m ::KIIJ::fij5f11UIIJ
fi1U7CrZIIJ~~:;K~~m. The reason is that the first statement comes with a com-
plete conclusion in a separate sentence, as does the second. This makes it unmanageable
to have a singlefo linking and marking both items.
47 In Wang's comments to Hex. g(1), 8(4), 9(5), 12(1), 15(1), 16(1), 26(1), g4(1), 44(1), 45(1), 49(1),
51(1), 52(1), 54(1), 57(1), 62(1), 6g(2), 64(1). In his ZYLL "Ming tuan" (1), "Mingyao" (1).
48 In Wang's comments to Hex.1(4) and in his ZYLL "Ming tuan" (1), "Mingyao" (1), "Bian wei"
(2). The definitions are better represented in the ZYLL because it gives a systematic analy-
sis of the structure of the Zhouyi. There are two items, where fo serves to introduce a
philosophical definition with a rhetorical question, :;K ~~{DJ m. Even in this case, I
would not think it is right to treatfo just as a marker of the beginning (although this is the
beginning of the entire ZYLL ), but would suggest a translation: "What, as a matter of gen-
eral principle, is a tuan?" The second item starts the section "Ming yao tongbian.'' It refers
to two such definitions. ::K 3Z ~{DJ m 8-'f~~m ~~{DJ m •IJ:!ffX?,Zr!T~m. "What, as
a matter of general principle, is a line statement? It is that which verbalizes a transforma-
tion; and what [as a matter of general principle] is a transformation? It is that which
brings about the true and the false.'' Lou Yulie, Wang Bijijiaoshi, 2.597.
49 In Wang's comments to Hex.1(g), 9(1), 15(1), 17(1), g2(1), 42(1), 45(1), 49(1), 51(1), 54(1), 57(1),
6g(l), 64(1).
so In Wang's ZYLL two are in the "Ming tuan," and one is in the "Ming yao.''
BUILDING BibcK oFciii:NEs'E.AR.GuME::Nl'A'rfoN: INITIAL Frr 7( 'SI
In the entire body ofWang Bi's surviving writings, I have not found a single
.... ~..item that did not fit the analysis given here. We are therefore entitled to reverse
the strategy. Instead of proving piece by piece that a given initialfo again is
used in this manner, we can now assume that all phrases in Wang Bi with ini-
tial}U must be reckoned as statements of a general [unknown and new] prin-
ciple or definition. And in a further step-one that is not taken here-this
wouia enable us to locate many if not most ofWang Bi's statements of general
principle or philosophical definition by looking for the phrases wi~h initialfo.
We have analysed initialfo as it occurs in a corpus of texts that IS narrowly
circumscribed in terms of time, genre and content. To situate initialfo in the
history of Chinese language and argument, I will now trace the history of this
feature.
A Sketch of the History of Initial.fit
Rhetorical devices have their history and particular context. While we may
safely say that initialfo as defined above is consistently used in Wang Bi's writ-
ings, this theoretically could be a stylistic idiosyncrasy. We will therefore go
backward through a selection of earlier texts to check this assumption and, if
possible, trace the path from a demonstrative use of initial fU to becoming the
marker of a phrase of general validity. Examples will be given in the notes.
Wang Fu's .:£{,4 (circa 76?-157? AD) Qiarifu tun )17(gRlJ is a well preserved
highly argumentative text. It has 127 sentences with initial.fit, distributed across
all thirty-six chapters but numbers g, 15, and 34· Of these only one occurrence
is demonstrative. 5l The proportion of these uses indicates that the use of initial
JU as a phrase status marker for a statement of general validity had stabilized
well before Wang Bi and that demonstrative use was becoming very rare in-
deed. In their great majority, the initialfo sentences in the QianfU tun are at the
beginning of an argument. This position, which we also find in Wang Bi's theo-
retical essays on the Zhouyi and the Laozi as well as in later non-narrative texts,
has led to the assumption that they are simply markers for a paragraph break.
Modern editors in turn have based their layout on this assumption.
51 Lao, D. C., Chen Fangzheng (Chen Fong Ching) ~Jit0IE comp., Qianfo lun zhuzi suoyin
>;g;:;Kg@~~%'7 [,cHANT database (Hong Kong: Commercial Press, 1995). Example for
initialfo as a phrase status marker: :;K ~;\.~7(0 Jf~~~ii' ~1&~;\.Z§ 7CZ
,c,,fu Jf~Zpff§Jt ~;\.Z~m (eh. 7) ('~sa general principle, the sages act as heaven's
mounts and the worthies as the translators of the sages. Therefore the words of the sages
are the mind of heaven and what the worthies explain is the meaning of the sages"). The
only example for demonstrative use will be found in eh. g5.

Going one step back to a similar type of text, Wang Chong's .:EJE (27--97?)
Lunheng glfij~, we find him using initialfo 172 times in the first 20 of his 85
chapters. Of these, 166 occurrences mark sentences of general validity, one is
demonstrative, 52 and there are six occurrences of phrases beginning withfo ru
shi :;k: 3<0:@, a highly idiosyncratic phrase beginning in Wang Chong that is as
rare in other pre-Han and Han texts (one or two occurrences) as it is frequent
in the Lunheng (44 occurrences). As demonstrative fit always refers to specific
entities such as persons and never to situations or processes and as all occur-
rences of :;k: 3<0:@ here present arguments, I suggest to read them as marldng
statements of general validity and translatefo ru shi as "As a matter of general
principle, [things] being like this ...."The use of initialfo in Wang Chong shows
the same high consistency as the later texts already studied.
Liu An's ;u~ (ea. 179-122 Be) Huainanzi >ti¥i-T (mid 2nd cent. Be) has
many occurrences of initialfo in chapters 1, 2, 6, 7, g, n, 12, and 20.53 All33 oc-
currences in chapter 1, one of the two chapters studied by Hans van Ess, con-
form to the definition. In many cases, the general principle there refers to a
principle underlying a process in nature rather than, as in the examples stud-
ied hitherto, in society. 14 of these 33 occurrences are followed by a 'that is why,'
t)Z: or ;@t)Z:, which make the link to a previously stated principle explicit. Ran-
dom checks in the other chapters confirmed the consistent use.
Jia Yi's Wm (2oo-168 Be) Xinshu iFJTil has a use of initialfo with eighty oc-
currences followed by a statement claiming general validity. There is no de-
monstrative use. To this has to be added his idiosyncratic formula gufo t)Z::;K
with twenty occurrences in these chapters. As these all contain statements of
general validity, I suggest translating them as "that is why as a matter of gen-
eral principle ...." Importantly, Jia Yi does not [yet] use initialfo regularly at the
beginning of an argument to ground it, but mostly at the end to sum it up.
Lii Buwei's §:fit (291-235 Be) Lilshi chunqiu §B(,lf;f)( (239 Be) already
has the same consistent use, as a check through the 38 occurrences of initial fo
in the first ten chapters shows. 54
52 Example for use as phrase status marker in the Lunheng; ;};( tc/f JJ!!!;$ *
OJ~!IItu :X:/f
*
ffll.Jl' OJ~Jlj~tt!, (As a general rule, if a comparison does not fit the matter at hand, one
cannot call it a metaphor and if a statement does not correspond to reality, one cannot
call it truthful). The only example for demonstrative use:;};( 31'f:E:t:f *frEi~Wtt!. (THIS
[previously mentioned] "King of Qin" is none other than Qin Shi Huangdi).
53 Lao, D.C., Chen Fangzheng (Chen Fong Ching) ~Jit:tJJE comp., Huainanzi zhuzi suoyin
)1£i¥f-T~::f:%:1§f, CHANT database (Hong Kong: Commercial Press, zoog). Example:
X. flxi\;fit !flliZY::tJ 1/F/\~ ~/for~ !~/f"OJ!~rr i>l~x:tfu JfHj't~~ (ch.1)
lli::t:r
CAs a general principle, the Dao covers heaven and carries earth, extends the four direc-
tions, opens up the eight extreme points, (its Jheight being unlimited, (its J depth unfath-
omable, it embraces heaven and earth and endows the shapeless").
54 Lao, D.C., Chen Fangzheng ( Chen Fong Ching) ~-*:1JJE comp., Liishi chunqiu zhuzi suoyin
BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU '!<: 59
The authors preceding Wang Bi, who have been studied here, all also make
frequent use of initialfan }L, another phrase status marker that so~etimes
even occurs in the same paragraph as initialfo. This has not been studied here
in detail so that a definition of its differences withfo has to be left to further
research.
I conclude that Wang Bi's use of initialfo was not idiosyncratic, but followed
a routine firmly established at least since the latter part of the third century BC.
Within this now established rhetorical practice, different authors still differed
in its particular use. Whether an author would push for philosophical insights
that might be radically new and even counterintuitive, restate more generally
accepted and perhaps less surprising principles, or would deal with principles
underlying natural processes, they were all operating within the same rhetori-
cal convention. As this is a practice and not a law, it cannot be excluded that
rare exceptions still persisted.
As may be expected, the use of initialfo is much less frequent in narrative
texts. Sima Qian's ~,~}I (circa14o-86 BC) Shiji Jt:§c is a work strong in narra-
tive but without a strong authorial voice stating matters of general principle. In
the direct speeches, however, which the Shiji amply attributes to or quotes
from historical actors, we find altogether about 44 occurrences of initial fo
marking statements of general principle or definition. 55 Here, we have two
phrase status markers, one, initialfo, marking the general precedent, and the
other, initialjin 9, marking the application to the case at hand, as is frequent
in legal documents. In hjs own statements, Sima Qian uses the feature only
once, saying "As a matter of general fact, I have no knowledge whatever of the
time before Shennong" :;k:_ t$J!lP)JrtJ 1§-:fffiE, a sentence that neatly describes
the historical time frame for which he could claim to have reliable evidence. 56
Pushing further back in time, we find that initialfo was used in two different
functions during the Warring States period, one being the status marker stud-
g 13()€ffX~::f:%:1§f, CHANT database (Hong Kong: Commercial Press, zoog). Example:
:X J;J1.i~1~~~c,,~:tr:kr}J.i"t:t:f J::ti!:§-1'f;L~ *~'L'fffl:fLJ/J.i"t:ti *Z
~1'1 tt!.
(eh. g.z) ('~sa general rule, to reach people's hearts through virtue and thus establish
merit and fame was frequent in olden times, but it has never happened that someone
failed to win the people's hearts but established merit and fame").
55 Example:;};( fifi;~~!ilf\ JWJEL:: ~f~1'f:==:!imm-~111T fffl~*>5m-ti!: Jl:tfli-T1iPJT ti,
flm-~t!:t. ('~a matter of principle, that Zhong and Li committed no crime whatever,
but [still] went to their death, while I now have committed three crimes before your
Majesty and still want to continue to live in this world, this is a case ofWu Zixu failing in
Wu"). Sima Qian '§],1~1 }!, Shiji )!:§[, (Beijing: Zhonghua shuju, 1973), 93.2635· Here, we
have two phrase status markers, one, initial }it, marking the general precedent, and the
other, initialjin ~. marking the application to the case at hand, as is frequent in legal
documents.
56 Ibid., 129.3253·
 . 6o
ied here, the other being that of a demonstrative. For either use no early exam-
ples have been found in in the standard sources, namely oracle bones, the
Shijing, Shangshu or Western Zhou bronzes. 57
An example for the use of initialfo as a demonstrative in the beginning of a
sentence will be found in the Zuozhuan. Reacting to the statement "Although
Chu has talents [enough], it is Jin that makes use of them," Zimu is quoted as
saying: r*El ::)(j!ilj~~fr!N.IJZ "Are they [Jin] alone in not going by clan and
in-laws [in assigning government jobs, R.W.]?"5B The commentator Du Yu HrJi
(222-284) confirms this demonstrative use with his comment:::)( ~~-y ''fore-
fers to [the state of] Jin." The sentence is part of a narrative dialogue and does
not state a principle. Indeed, in all narrative sentences that start with an initial
fo, its meaning is demonstrative. The dividing line between the two kinds of
initialfo is neither graphic nor phonetic, but whether the sentence that follows
is narrative or states a principle. If the use is demonstrative, this also presup-
poses that the object has been mentioned previously. The use as a demonstra-
tive was never widespread and ended by the late third century BC. This did not
affect the use offo ::)( as a demonstrative pronoun within a phrase as in the
Zuozhuan statement: B~~:k:.0t.~¥Jk.f.&ij6{:f~$ffij~Jt~.JHEB "Because the
luminous lord considered THIS Gongsun Duan as capable, he entrusted this
affair to him and gave him an enfeoffment of Zhoutian." 59 Examples of this
kind will be found as late as the Ming and Qing.
The use of initialfo as a status marker is much more widespread already in
the pre-Qin period. Because of the difficulty of dating pre-Qin texts with any
precision, it will be possible to determine that initialfo was used as a status
marker in argumentative texts of this period, but no clear chronology can be
given. Some early texts do make consistent and some even frequent use of the
device. They might have been instrumental in establishing the routine. The six
cases where it is used in the Xiaojing all correspond to this pattern. An exam-
ple: rEI :k:. # t~~:$:ili ~~?JTE8~ili "The Master said: 'As a matter of gen-
eral principle, filial piety is the basis of receipt/capacity and it is that on the
basis of which education comes about."' 60 The Sunzi with its highly senten-
tious offerings of a systematic theoretical treatment of war has ten
occurrences, all of them fully consistent with the later pattern. 61 The most
57 WangWeifeng, "Fu," 67.
58 Zuozhuan, Duke Xiang ~0 26th year. In Ruan Yuan ~3Cjj'; ed., Chongkan Song ban
Zuozhuan zhushu: fu jiaokan ji ;!:**t&:h:ff5.±)(,1t: Jlftt'l(:WJ§c (Shanghai: Jinzhang
shuju, 1932), 635.
59 Zuozhuan, Duke Zhao Bjg-L}, 7th year, ibid., 763.
6o Ruan Yuan ~Jtjj'; ed., Xiaojing, Chongkan Song ban Xiaojing zhushu:fu jiaokan ji ;!:**
t&:~~5.±iffit: Jlft~WJ§c (Shanghai:Jinzhang shuju, 1932), 10.
61 Sunzi bingfa :!'-*-TA$ (Changsha: Yue Lu shushe, 1993).
/1. BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU ~
extensive use of initialfo in pre-Han texts will be found in the Guanzi "gr,
where I count no less than 242 occurrences that fit the pattern. They are espe-
cially frequent in chapters four (13), six (20 ), nine (45), fifteen (66), seventeen
(35), twenty-one (12), and twenty-three (15). 62 •
The Laozi in the reconstructed Wang Bi edition shows 27 occurrences of im-
tialfo, which for such a short text is a large number. Of these nine have the
formulafo wei ::Kotf (2, 8, 15, 15, 22, 70, 71, 72, 75); 63 in eight cases this formula is
followed by a gu ti)( or shiyi ~tJ in the beginning of the next segment, giving a
meaning "and it is a fact that only because of X the result will beY," while five
more just have fo ;K followed by gu ti)( or shiyi ~tJ in the next segment. The
remaining occurrences such asfo hegu ;K fBJti)( in zhang 50 might be leftovers
from the demonstrative use of initialfo that predates its consolidation into a
standard rhetorical device. The Laozi makes use of initialfo to stake general
claims about counterintuitive dynamics. "It is exactly because he [the Sage]
does not install [himself in these particular achievements] that they do not
disappear." ;K OJE::f"@ ~J;J::f"-'$;.64 Only in one single case (zhang 38) does it
usefu for a general definitional statement.
It is a general principle [,however,] that ritual is [the result of the] wearing thin
of truthfulness and credibility and [thus] the beginning of [social] chaos; while
foreknowledge is [the result) of the Way's becoming an [external] ornament and
[thus) the beginning of stupidity [=inept tyranny].
The use of initialfu is highly consistent within the Laozi, even if it differs idio-
syncratically from that of other texts. It marks statements of general validity.
The Lun:yu quotes Confucius several times with initialfu phrases, and with
one single exception (174) they fit the notion ofju as a status marker. These
uses range from general definitions such as ::K e~ ...,
''As a matter of princi-
ple, someone with humaneness is ... " (6.30) ;K }¥~ ... , ''As a matter of principle
someone with attainment it is ... " (12.20) ;K :Er~@~ ... "As a matter of prin-
ciple the gentleman when in mourning will ... " (17.21) to emphatic markers to
set off his own practice from that of others as in ;K ~~U/FB~j ... (14.29). 65 With
62 Lao, D. C., Chen Fangzheng (Chen Fong Ching) ~Jf!JJIE comp., Guanzi zhuzi suoyin it-T
~"F~i§ I (Hong Kong: Commercial Press, 2001).
63 The two Mawangdui texts share this feature with one exception (71). The only overlap
with the Guodian Laozi is eh. 2, where Guodian A also has thefo.
64 Laozi 2.5. The formula is in both the Guodian A and the MWD B MSS.
65 Yang Bojun thf8~. Lunyu yi zhu iilflliit'fii~5.± (Beijing: Zhonghua shuju, zoo6), 72, 146, 212,
175·

altogether ten such occurrences, the Lunyu reflects a use of initialfo that is
fairly consistent. Given the strong elements of spoken language in the Lunyu,
the use of initialfo as a status marker might have its origin in oral communica-
tion. While the Zhouyi !EO~ itself does not have any initialfo, two of its philo-
sophical 'wings' do. The "Xici" (6) and "Wenyan" (2) segments both use initial
fo in a routine manner in the sense outlined here without any deviation, which
signals that these terse texts were drawing on an already established routine,
The bamboo texts have by and large confirmed this sketch.
In these occurrences, features occur which develop more complex rhetori-
cal patterns. One such pattern is the confrontation of a general principle intro-
duced by fo with a particular act deviating from it introduced by )in ~;66
another is the deduction of an 'inevitable' consequence flowing from the prin-
ciple enunciated earlier, marked by bi ~~; 67 and a third the application of the
principle under consideration to the situation at hand with a gu tf;)(, 'that is
why ....'
Evidently, this is only a loose sketch made largely possible by the availability
of full-text databases (CHANT, Siku quanshu, and Academia Sinica). It does
show, however, that patterns can be discovered and that one might even use
the innocent initialfo to test textual consistency and linkages between differ-
ent versions of a given text. Initial fo fulfills one important feature familiar
from authenticity studies in art history, namely that it is not a strong (termino-
logical or argumentative) identifier that would be picked up by someone going
for creative text enrichment or homogenization, but is rather a silent and in-
nocuous device. At the same time its function as a phrase status marker offers
an altogether different research option.
Conclusions
The study suggests: Initialfo is written :J( in a loan from the homoiophonous
fo :J( 'adult male: It is differentiated by context fromfo :J( 'adult male;' by
position in the phrase from a homoiophonous wordfo 'is it not so?' that is also
written :J( and terminates a phrase; and from demonstrative initialfo through
the type of sentence it starts, namely a narrative phrase in the case of demon-
strative initialfo, and an abstract subject with a definitional or argumentative
phrase in the case of non-demonstrative initialfo.
Scholars have speculated that what I called the status marker might have
developed out of the demonstrative, but the evidence seems weak. All uses of
66 Examples are Lunyu 17.21, Guanzi 5.52.30 and 5·53·9·
67 An example is Guanzi 4·32·5·
A BUILDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU "#: 6g
the character :J( for particles developed during the Warring States period. Al-
ready during this period, initialfo followed by a statement of principle was
used much more frequently than demonstrative initial fo with a narrative
phrase following. By the late third cent~_?' BC, the non-demonstrativ: u~e ~as
the lone survivor, and words such as bt 111l. had replaced demonstrative Initial
fo. Only rare examples have been found for the use of demonstrative initialfo
even in strongly narrative historical texts. Within sentences, however, fo :j(
continued to be used as a demonstrative, as this use was not burdened with
ambiguity.
Some Warring States texts already made predominant or exclusive use of
the phrase status marker function ("Xici" and "Wenyan" in Zhouyi; Xiaojing;
Sunzi; Laozi; Guanzi). Their status and the presumed link to Confucius in the
first three may have been instrumental in popularizing this usage. The devel-
opment of initialfo into a phrase status marker for statements of abstract prin-
ciple must be read in the context of the shift during the Warring States period
to thinldng in abstract categories and the need to develop a rhetoric of philo-
sophical argument to express, structure, and hierarchize it.
Departing from the imported prioritization of the 'classics' or the 'classical'
period in the study of grammar and rhetoric, the present study first looked at
the rhetorical feature of this phrase status marker in its most developed and
pervasive form by taking a mid-third century AD philosophical text (Wang Bi's
Zhouyi Commentary and "Liieli"). In an effort to avoid the standard mistake of
including only examples that fit the argument it presented a randomized sam-
ple. Hypotheses were developed through the analysis of individual occurrenc-
es, and tested against the random sample.
The result showed a highly consistent use of the rhetorical feature of initial
Ju. It sel'Ves the overall function of a sentence status marker for statements of
abstract principle. It refers to the entire subsequent sentence, which might
consist of two parallel segments. These statements of principle are stated, not
argued; their claim to validity is mostly agreement with the insights of the Sag-
es as contained in their written bequests with which the readers are supposed
to be familiar. (In Buddhist writings, this would refer to Buddhist "classics").
The proof is often contained in the convincing applicability of the principle
stated in the fo sentence for deciphering the coding underlying these texts or
for understanding real-world dynamics. Such proof is often highlighted by
separate sentences that start with an expression such as 'that is why' gu tli)z:, or
by linked phrases that start with a ze ~U 'as a consequence' and might contain
abi &;, ('bynecessity') or a similarword. 68 The claim to validity might be coun-
terintuitive, but it is implied that it should be convincing not just
68 This link has been highlighted by Wang Weifeng, "Fu."

in the contexts of the beliefs of a school or a time. The claim is a rational one.
Acceptance does not hinge on religious belief or submission to authority. Giv-
en these features,fo phrases are concentrated in non-narrative argumentative
genres.
While I would claim that the more formal features (such as status marker for
statement of principle, non-narrative phrase and abstract subject, and concen-
tration in non-narrative genres) are shared by all occurrences, there is some
bandwidth in the content definition that needs to be further explored. The
uses of this initialfo in texts after the third century AD have not been investi-
gated in this study. Anecdotal evidence, which includes Chinese Buddhist writ-
ings and commentaries, suggests that the use of initialfo as a phrase status
marker remained a staple in the rhetorical toolbox of Chinese argumentation
until the shift to the modern vernacular.
The study has a philosophical implication by enhancing the precision in the
understanding of arguments as it allows us to properly hierarchize different
philosophical utterances. Its definition of initialfo furthermore can be used for
very practical textual diagnosis: namely the location of such statements of
principle so as to be able to insert them into a topical map of abstract princi-
ples that might be useful for studies of an individual author, of a group, or of
broader questions in the history of ideas.
The fact that initial.fo marks statements for which general validity is claimed
does not imply that all such statements start with initialfo. It will require fur-
ther research to ascertain whether statements thus marked share certain fea-
tures that are not found in other statements with a claim to general validity, or
whether use of initial fo is discretionary. While this means that no claim is
made that all statements of this kind will be located (for the period after the
stabilization of this use) by searching for initialfo, such a search will still yield
substantial results of value for intellectual historians.
Apart from having to be subjected to further randomized tests (and possibly
refinements of the analysis), the study suggests further explorations. One evi-
dent question would be whether the development of initial fo is part of a
broader development of rhetorical tools to express abstract arguments. Initial
fan fL is a case in point that has been mentioned. Interlocking parallel style
(IPS) developed during the same time horizon as initialfo into such a tool. It
came with its own status marker for a statement of abstract validity, namely
the (normally) non-parallel concluding phrase after a string of interlocked par-
allel phrases going each through one half each of a totality (heaven/earth;
fame/wealth; social standing/personal security; objects/processes, etc.).6 9
6g It will be noted that many modern authors writing on the Laozihave rightly; yet spontane-
ously-and without much understanding of the underlying interlocking parallel style
pattern-selected these phrases as the mainstay of their analysis.
LDING BLOCK OF CHINESE ARGUMENTATION: INITIAL FU '*':
the innocently-lookinggu i!& or shiyi ~PJ might offer some surprises and
would need more careful study.
.Chinese writers of philosophical texts used a wide variety of tools that were
. art purely structural and formal and in part explicitly rhetorical, to reduce
Ill p
ambiguity in the presentation of their arguments wh'l '
1 e staymgWI 'th'm th e co~-
ntions of terse and pointed language. One of their greatest challenges was m
;:ducing ambiguity in their use of abstract concepts. Their standard tech-
niques such as organizing arguments along mutually-defining ant~nyms have
been supplemented by the sophisticated and consistent use of s1lent (struc-
tural) as well as articulated tools to clearly mark and structure statements. Ini-
tialfo ;j( is one of these tools.
References
Baxter, William H., and Laurent Sagart. Old Chinese: a new reconstruction. New York:
Oxford University Press, 2014.
- . "The Baxter-Sagart reconstruction of Old Chinese (Version 1.1, 20 September
2014 )." http:/ /ocbaxtersagart.lsait.lsa.umich.edu/ (last access o6/o3/2015).
pongfangshudian *1Ji'if]2::f, ed. Guoyu da cidian ~a'l!i:k~$. Taipei: Dongfang shu-
dian, 1972.
Graham, A. C. "The Final Particle 'Fwu' ;;J(." Bulletin ofthe School of Oriental and African
Studies 17-1 (1955): 120-1;32.
~."The Classical Chinese Topic-Marker fu ;;J(." Bulletin of the School of Oriental
and African Studies 35.1 (1972): Ss-no.
Harbsmeier, Christoph. "Chinese Rhetoric in Comparative Perspective." Unpublished
manuscript, 2002.
Harbsmeier, Christoph, et al. "Thesaurus Linguae Sericae (TLS):' http:/ /tls.uni-hd.de/
(last access 20/n/2014).
Lau, D. C., and Chen Fangzheng (Chen Fong Ching) ~t1JI£, eds. Guanzi zhuzi suoyin
tg-T~*~i§ J. Hong Kong: Commercial Press, 2001.
- - , eds. Huainanzizhuzi suoyin )i£wg-T~*~i§ J, Hong Kong: Commercial Press,
2003.
- - , eds. Lilshi chunqiuzhuzi suoyin g [X,lffk~*~i§ [.Hong Kong: Commercial
Press, 2003.
- - , eds. Qianfulunzhuzisuoyin 5!;;J(~Ifll~*~i§ [.Hong Kong: Commercial Press,
1995·
Lou Yulie i'tf;sr!t ed. Wang Bijijiaoshi _:f.585~~~~. Beijing: Zhonghua shuju, 1999·
Luo Zhufeng mt-t J!!R, ed. Hanyu da cidian ~a'/!i;kg~$. Shanghai: Shanghai cishu chu-
banshe, 1986-1994.
Lynn, Richard. I Ching. New York: Columbia University Press, 1994·
 MaJianzhong '~ fr!,'[t',. Ma shiwentongjiaozhu ,~ f\:.)(}ji'3()1:. Commentated by Zhang
CHAPTER 2
Xichen 1T/H~l~. Beijing: Zhonghua shuju, 1988.
Pei Xuehai ~!'.~· Gushu xuzijishi tfiili@:*~~. Beijing: Zhonghua shuju, 1954.
Ruan :uan fmn, ed~ Chongkan Song ban Zuozhuan zhushu: fo jiaokan ji ]!.Bi~5Kf&.:£
Beyond Parallelism: A Rethinking of Patterns of
f•>1:1F.Ift: flfttxWJs2. Shanghai:Jinzhang shuju, 1932. Coordination and Subordination in Chinese
--~-,:d. Chongkan Song banXiaojing zhushu:fojiaokan ji 11H~5Kf&~~).!ih'ft: fl("f
fxWJs2. Vol. n. Shanghai:Jinzhang shuju, 1932. Expository Prose
- , - , ed. Z~ouY}_ jiaokan ji fED ~;f)(WJ§2. In Shisan jing zhushu: fo jiaokan ji +--=:~
Andrew H. Plaks
j~IFJit: flft~~WJs2, vo~.1, edited by Ruan Yuan. Beijing: Zhonghua shuju, 1987: 5-108.
Schmdler, B. Grammatical Notes n: :;X.." Asia Major, N.S. III.z (1953): 16z-168.
Schoenhals, Michael. Doing Things with Words in Chinese Politics: Five Studies. China
Among the papers collected in this volume dedicated to logical and rhetorical
Research Monograph 41. Berkeley: Institute of Chinese Studies, University of
California, 1976. patterns of composition in Chinese texts, under the broad heading 'literary
Shisan_!:
jing cidian bianzuan weiyuanhui +=£:(((Slit .f!I:J.£:15 ~ =*' I§~ Sh tsan;mgct
· .. 'dtan
· forms of argument; a substantial number (those dealing with 'palindromic
~"tt:l5iJ-t_,.,"vii!J~>e-""s·
+ .=:"!m!lt!-. Xi'an: Shaanxi renmin chubanshe, zooz. structures; 'topological clusters; 'two and three-step exposition,' 'semiotic
Sima Qian OJ,~}JJi. ShijitJ::.§2. Beijing: Zhonghua shuju, 1973 . webs,' 'tetra-syllabic rhythms' and the like) seem to focus on various techniques
Song Lian 5K5J! (1310-1381).Pianhai leibian ~~;J!Ji~i. Shanghai: Shanghai guji chuban- of verbal manipulation used to bind separate prose units into more complex
she, 1995-1999. discursive structures.1 Most of these studies appear to take as more or less self-
Sunzi bingfa f~-T Ji:-5"*. Changsha: Yue Lu shushe, 1993. evident the observation that the fundamental grid of classical Chinese exposi-
Van Ess, Hans. ''Argument and Persuasion in the First Chapter of Huainanzi and its Use tory prose tends to take the form of sets of parallel utterances. Typically, these
of Particles." Oriens Extremus 45 (zoos): 255-270. verbal strings are reeled out in sequences of equal numbers of characters,
Wagner, Rudolf. The Craft ofa Chinese Commentator: Wang Bi on theLaozi. Albany: SUNY forming paired units that are to one extent or another isomorphic in their
Press, zooo. grammatical or syntactic function. 2 Such parallel constructions are most visi-
- - - . Language, Ontology and Political Philosophy in China. Wang El's Scholarly ble-in fact they form the aesthetic core-in lilshi ('regulated verse') and oth-
Exploration of the Dark (Xuanxue ). Albany: SUNY Press, 2003 . er major modes of classical Chinese poetry, but they are by no means
--.-.·A Chinese Reading of the Daod~ing: Wang El's commentary on the Laozi with absent-albeit in a number of different forms-in most genres of guwen ('an-
crtttcal text and translation. Albany: SUNY Press, zoo3. cient-style prose') composition. Even in the primarily non-classical mode of
Wang Haifen .:E -$i::tt G rr · • -+ :'1i':-'-
''-'J-*• u Hanyu xuct• ctdtan _,_ _,_
n /~sftiilll:siiJsliJ!lt!-. Beijing: Beijing Daxue traditional fiction, paired lines continue to constitute a very prominent com-
chubanshe, 1996. positional feature, especially conspicuous in the conventional use of parallel
Wang Weifeng .:E~J-. "'Fu' de tongshe lianjie gongneng lun" :;X_!¥J~Jt:JI~t~JJJfj5~l1l/. couplets as chapter titles, reflecting a loose pairing of the main narrative units
GuHanyuyanjiu tfi5l§:S:EJf~ 83.2 (2009): 63-68. comprising a given chapter. When this paradigmatic use of parallel sequences
Xing Bing lf~~. Sub-commentary in Xiaojing zhushu ~~ti!Ff,t. In Ruan Yuan ~)t:J[; is developed to such a degree that it comes to constitute the primary feature of
ed., Shisw}jing zhushu
Xu Kuanmrl ,tr::~~ Hu ·
+-£~).!iht Beijing: Zhonghua shuju, 1987: 2539-2561.
'll:""".:z..L...~!?E!
a part or the whole of a given piece of prose or verse, we are justified in apply-
oJ• 1:1 I ~=s · amanzt· quanyt· lttl=f'J J ±S"P· Guiyang: Guizhou renmin chu-
banshe, 1995.
Xu Weijian filff¥fr!, ed. ShangguHanyu cidian J:tfi)J;§:S:§iiJ!lt!-. Changchun:Jilin wen-
shi chubanshe, 1998. 1 The terms applied to such units vary widely, including such things as 'colon,' 'lemma' and
Yang Bojun mf~ ~. Lunyu yi zhu ~l1lJ§ft~)1:. Beijing: Zhonghua shuju, zoo6. others, depending upon their scale and function in a given analysis.
z Quantitative parallelism in Chinese prose was later regularized in various modes and genres,
Zhou. Fa~ao fED/"*~. Zhongguo gudai yufa tp ~ tfi {-\§:S: )"*, Chengdai pian fj"ij. {-\~.
for example in what came to be known as 'four-six prose' (siliuwen ), and in later periods pro-
T~1pe1: The Institute of History and Philology, 1959; reprint, Beijing: Zhonghua shu- duced parallel strings of extreme length in the so-called bagu ('four-legged') form used in
dian, 1990.
examination essays, and in other genres of prose composition.